Collaboration between ESS and DESY • Workshop on Intelligent Control Systems

ESS is already involved in a pilot study, ESS Control System Data Lab, together with industry and academia, focusing on alarm management. ESS is now entering a new collaboration with the DESY research facility.

In September, the ESS DESY Intelligent Control System Workshop was arranged to discuss collaboration on artificial intelligence in control systems for accelerator-based research facilities. Representatives from the facilities and from industry and academia held discussions to identify the areas of technology where collaboration is suitable and the objective of the collaboration.

A reference group has been set up with industry, including big companies such as ABB, Hitachi, Tetra Pak and Perstorp. The reference group also met in September, and held interesting discussions about longer projects on, for example, complex systems and open data.

Anna Hall, Director of Big Science Sweden, participated in the workshop.

“This is the start of a long-term collaboration that will benefit both research facilities, industry and academia. By starting collaboration projects in northern Europe, we’re building a knowledge corridor that will strengthen the entire region in terms of innovation and research.

“The facilities’ experiences of handling large quantities of data, advanced control systems, and open data are important for industry and the autonomous factories of the future, with exciting new technology in fields such as AI and machine learning.”

Karin Rathsman, accelerator physicist at ESS and Project Manager of Control System Machine Learning (CSML) at ESS, hosted the workshop together with Annika Eichler, Senior Researcher at DESY. ESS has positive experience of collaboration in the ESS Control System Data Lab, which discusses, for example, how to share knowledge about open data. The workshop in September considered a possible collaboration between ESS and DESY, and what would be needed to get things under way.

Karin Rathsman is looking forward to the collaboration.

“As in the work with ESS Data Lab, we’ll be focusing on control systems and alarms. We’re starting from the bottom, linking together people who work with the same things at our respective facilities. It’s good to start with something concrete, then we can work closely and develop methods together. We find out what we can each contribute, and build up a collaboration that adds value for everyone.”

As a new facility, ESS has been able to build up an efficient control system from the start. DESY, which has been in operation for over 50 years, has long experience, much knowledge to share, and is constantly developing.

Annika Eichler from DESY shares Karin Rathsman’s view on how successful collaboration can get under way.

“We’re now building a collaboration platform. In the short perspective, it’s important that we start talking with one another, and that we find tangible, common areas to work with. The workshop included presentations and constructive small-group discussions. Now we’ll see how the small groups can continue working. There’s a positive feeling about the collaboration, which will interest more parts of our organisation.”

A deeper collaboration between ESS and DESY has a good basis for success. The facilities are close, not just geographically, but also culturally. They are comparable in size. Both are interested in developing the initial areas selected for the collaboration, data sharing and alarm management. 

Jan Eric Larsson is President and CEO of GoalArt, a company involved in the pilot project ESS Control System Data Lab and in contacts with DESY.

“As alarm specialists, we can contribute knowledge and discussions regarding alarm philosophy, which data can be registered, and how data is stored. Large facilities with many sub-systems face a challenge in creating a common philosophy for control rooms, names of signals, etc.

“ESS and DESY can inspire each other to new development projects, they can share expertise, and work together. The workshop was a good start for making contacts and learning a bit more about what is currently being done in each facility.”

Industry, academia, and research facilities have a common interest in increasing collaboration on open data and open innovation, to drive development in AI and machine learning. Per Runeson, Professor in Software Engineering at Lund University Faculty of Engineering (LTH), has led a feasibility study on open data for machine learning, and is also involved in the pilot project with DESY.

“Industry and research facilities work with related issues, one of which is machine learning. Experiences and findings from ESS can spread to Swedish industry and vice versa. One fundamental issue is management of data and open data. At management level, there can be a fear of open data, but at operative level, researchers and engineers have a more positive attitude to data sharing. Here, it’s about finding a balance, and the collaboration between ESS and DESY can help with this; instead of building walls, seeing the value in greater exchanges and deeper collaboration.”

The collaboration is now continuing with more in-depth discussions, both within and between the research facilities.  The workshop on Intelligent Control Systems will be followed up with the reference group.

Anna Hall, Big Science Sweden, is also positive about the collaboration.

“Long-term collaboration requires curiosity, perseverance, and commitment from everyone involved. Our introductory discussions show clearly that there is commitment, and that this collaboration concerns sharing experiences and disseminating knowledge that strengthens both industry and the ESS and DESY research facilities.

“We see it as an exciting building block in the work to drive technology development and innovation. This is an important part of our work on knowledge transfer, and we are looking at tangible opportunities for larger projects to drive development.”